We are putting the finishing touches on our plans for the rest of the year. Next week, we are launching a new event format, The TRB Audience Summit, which will bring together top leaders in the media industry on Sept 17 in NYC for a day of interactive discussions and case studies around building audience-focused media companies. This is the most ambitious event we’ve done, and I’m excited to share why we believe it will be worth attending when I know there are many events. Get in touch to discuss partnership opportunities: [email protected].

My conversation with USA Today Co CEO Mike Reed ignited a fair bit of discussion for his bold statement that he’s considering delisting the publisher altogether from Google by blocking its search crawler. You can see TRB Conversations on YouTube, and we’re beginning to publish these conversations on The Rebooting’s podcast feed as well. Thanks to EX.CO for supporting this series.

Be sure to check out my conversation with Vox Media’s head of e-commerce and affiliate Camilla Cho and NucleusLinks CEO Cornelius Frey. We discussed how Vox has managed to grow its affiliate revenue despite a 25%-plus traffic decline to its commerce pages. See the replay. 

How to fix CTV revenue loss 

Programmatic CTV spend is rising, but revenue isn't keeping pace for most media owners. Nearly 80% believe they're leaving money on the table — the culprit isn't demand, it's infrastructure built for display, not CTV. EX.CO's new guide breaks down the three hidden revenue leaks costing you yield (signal loss, static floors, and outdated auction architecture), plus a four-step audit you can run in two weeks. See how one publisher grew CTV revenue 33% in 30 days with no added headcount. Get the data, the framework, and the fix.

The AI content paradox

Blame Satya Nadella for introducing Jevon’s Paradox to the AI babble. The Microsoft chief used the economic principle that holds that as tech makes a resource cheaper and more efficient, consumption increases. It’s regularly flogged by podcasters who point out that the predicted jobs apocalypse – always marketed by AI people, not economists – never came to be.

The AI content paradox is that as AI makes it cheap and easy to create all manner of creative output, the value content on top of the S curve will go higher. AI has not eliminated the need for humans to write or design. And I don’t think it’s going to happen.

Professional provocateur Marc Andreessen promised that AI will soon iron out the final "idiosyncrasies" that cause it to be widely loathed. Of course, as Chapin Clark pointed out in response, idiosyncrasies are markers of humanity, and humans are adaptable so they’ll just come up with new idiosyncrasies. Andreessen deleted his X post. You can just delete things. 

Meanwhile, Airbnb CEO Brian Chesky was met with ridicule for rolling out a clearly AI-generated X thread about crypto of all things. Chesky deleted the thread because, again, you can just delete things. 

There’s still a social cost to using AI to express yourself or your ideas. It’s embarrassing, even somewhat offensive, to outsource that. Serious people don’t behave this way.

I’ve heard all the arguments of people like Andreessen that this social cost is temporary. Look at this clever AI-generated video that brought Maggie Thatcher back from the dead. And so we’ll have a flood of this kind of AI slop. Some will be good or at least fine, much of it will be bad. Little of it will have much lasting impact. 

It will not kill many categories. I find myself numb to AI content at this point. I know the tells, and my knee-jerk reaction is to discount an AI-generated piece of communication. So far, all those AI-generated podcast guest pitches are about 0 for 600 in booking a guest on The Rebooting Show. To the executives using these services: There is also a social cost for this sweaty behavior.

My hypothesis is that companies will see diminishing returns trading humans for tokens in many of these fields. AI as a creation tool is clearly flawed. 

You’d think AI at this point would have killed aggregation. If AI is going to cure cancer, it should be able to aggregate. And to be sure, there are so many AI-generated newsletters out there. There’s a lot of talk in newsletter circles of declining engagement. I suspect that’s mostly a result of AI, not replacing newsletters but being a too hard to resist shortcut for the growth hacker crowd that piled into the newsletter boom of 2021-2024.

Peter Kafka asked New York Times executive editor Joe Kahn for his favorite newsletters. He named News Items, an aggregation newsletter put together by former Dow Jones exec John Ellis. John has an idiosyncratic method that involves Ben Franklin-like sleep patterns to unearth a remarkable diversity of information nuggets, neatly packaged so people do not need to click to read the original source. (All’s fair in love and aggregation.) News Items might have one of the more influential subscriber lists in media. Many influential and powerful people are fans of it.

It’s the same story with Feed Me. Yes, Emily does original reporting, but the heart of the newsletter is aggregation. 1440 has shown that there’s a continued market for human aggregation with a daily newsletter that now has nearly 5 million subscribers. I regularly ask Tim about why AI isn’t going to eliminate 1440’s value proposition. He believes curious humans want smart curation from other curious humans. Tim’s talking his own book, but I agree with him.

A lot of that goes back to friction. AI is positioned as something beyond the regular tech advancement. But it’s in line with the typical promise of removing friction. And the paradox that when you remove the friction of effort in creation, you actually raise the demand and therefore value of effortful content. News Items but done by AI is a less satisfying and successful product.

This will be repeated across many fields. A designer friend of mine who is very talented told me they are applying for a job in a bookstore because there’s no design work. That will not last. Just as AI writing makes many people’s skin crawl, so too does AI design. The reason is the importance of good design is that it signals you give a shit. 

The market will correct in time. The dream of agentic advertising is about efficiency and cutting out middlemen. Then what? My theory is we will see a return to true digital marketing creativity. Look at the rosé brand that created an entire alternative universe to stand out rather than trying to compete through automated ad systems. I expect a resurgence of analog media formats like niche magazines because to put together a magazine as opposed to a newsletter is a lot of effort. 

Even if AI can mimic the appearance of craftsmanship, it won’t be able to accrue the value people will always place on effort. 

Thanks for reading. Send me a note with your feedback by hitting reply.

Keep Reading